DigiNotar CA breach widens, Microsoft, Dutch government take action

The extent of the breach at Dutch certificate authority, DigiNotar, has broadened this weekafter an audit report analyzing DigiNotars servers released by the Dutch government showed majorsecurity lapses in the firms various CA servers.

Organizations need to understand what they should do if their SSL VPN would break for theirusers or if their e-commerce system would falter with their customers.

Chester Wisniewski, senior security advisor, Sophos LLC

The report, prepared by IT security firm Fox-IT, found the DigiNotar network had been severelybreached compromising more than two dozen CA servers. The extent of the damage increasedsubstantially, with evidence of CA servers that issued hundreds of signed rogue certificatesagainst 20 different domains. 

Some experts said the seriousness of the breach shines a light on the problems that plague thecertificate system. Chester Wisniewski, a senior security advisor at UK-based security vendorSophos LLC, said enterprise CISOs need to understand how their organization uses SSL certificatesand come up with a contingency plan if the certificate provider is breached. In addition to SSL usein browsers to verify the authenticity of a website, many enterprises use digital certificates toauthenticate users for SSL VPNs and email servers.

Organizations need to understand what they should do if their SSL VPN would break for theirusers or if their e-commerce system would falter with their customers, Wisniewski said. Askyourself: Is there an alternative plan?

Organizations can obtain certificates from multiple certificate authorities to have a back-upplan for website validation if a CA is breached, he said. Alternatives to the current systemare being tested, but until Google, Microsoft and Mozilla begin to support alternative authenticityvalidation systems, the system is unlikely to change. The Fox-IT report has prompted those browsermakers to blacklist DigiNotar certificates.

Microsoft updated its security advisory Tuesday, pushing out an automatic update to allsupported versions of Windows, revoking the trustin DigiNotar root certificates. The company said it made the move to protect users of InternetExplorer from man-in-the-middle attacks. Rogue digital certificates also enable attackers tospoof content and perform phishing attacks.

Weve deemed all DigiNotar certificates to be untrustworthy and have moved them to theUntrusted Certificate Store, wrote Dave Forstrom, director of Microsoft Trustworthy Computing inthe Microsoft Security Response Center blog. We recognize this issue as an industry problem, andwe have been actively collaborating with certificate authorities, governments and software vendorsto help protect our mutual customers.

Microsoft is waiting a week before rolling out an automatic update to users in the Netherlands.Mozilla and Google have taken similar steps to block the roguedigital certificates.

This is not a temporary suspension, it is a complete removal from our trusted root program,wrote Jonathan Nightingale, director of Firefox engineering in the Mozilla Security blog.Completerevocation of trust is a decision we treat with careful consideration, and employ as a lastresort.

Nightingale said the complete removal ofthe trusted root was taken because the scope of the breach remains unknown.  In addition,DigiNotar revoked fraudulent certificates without notifying Mozilla.

In an update issued Sept. 3, Google said it is rejectingall Certificate Authorities operated by DigiNotar. We encourage DigiNotar to provide acomplete analysis of the situation, wrote Heather Adkins, Googles information securitymanager.

The Fox-IT report, which was released by the Dutch government, found serious problems withDigiNotars network. 

All CA servers were members of one Windows domain, which made it possible to access them allusing one obtained user/password combination, according to the DigiNotarbreach report, which was made available on the Dutch government website Rijksoverheid. Thepassword was not very strong and could easily be brute-forced.

In addition, the audit investigation found outdated software installed on the DigiNotar publicWeb servers. No antivirus protection was present on the investigated servers, Fox-IT said.

Traces of hacker activity, believed to have emanated from Iran, began June 19 and lasted untilJuly 22. The attackers issued hundreds of rogue certificates, including an SSL certificate forGoogle, Skype, Mozilla add-ons, Microsoft update and others.

DigiNotar revoked the certificates and has added security measures on infrastructure, systemmonitoring and Online Certificate Status Protocol (OCSP) validation to identify the use of roguecertificates and prevent further attacks.

The security measures may have been too late. The report suggests the attackers used the stolenGoogle SSL certificate to snoop on users of Gmail in Iran. Log data analysis found 300,000 uniqueIP requests to Google.com with 99% originating from Iran, according to the report.

In a statement, VASCO Data Security International Inc., which owns DigiNotar, said it wouldfully cooperate with authorities and welcomed a fullreview of its systems by the Dutch government. As part of its proposal, VASCO invites the DutchGovernment to send staff to work together to jointly assess and remedy the problem.

CIA, Mossad, Also Targeted in Massive DigiNotar Cert Breach

The list of fraudulent certificates obtained by hackers who breached a Dutch certificate authority has grown to more than 500 and includes certificates for domains owned by three intelligence agencies: the CIA, Israel’s Mossad and the UK’s MI6.

DigiNotar, which is owned by Illinois-based Vasco Data Security, also lacked basic security safeguards, such as strong passwords, anti-virus protection, up-to-date software patches, according to a third-party audit conducted by security firm Fox-IT in the Netherlands, released Monday.

DigiNotar acknowledged last week that it became aware it had been breached on July 19, though it has never disclosed how long the hackers were inside its network before they were discovered.

DigiNotar is one of numerous firms around the world that generate security certificates for internet entities. The certificates authenticate web pages using the Secure Socket Layer protocol so that users can trust that their encrypted communication is going to the correct location. Anyone who manages to steal a certificate – such as criminals or government agents – can impersonate a legitimate site to steal log-in credentials and read a user’s communications.

Since news of the DigiNotar breach broke last week, the list of fraudulent certificates the hackers obtained has grown to at least 531, all of which have been disclosed by parties other than DigiNotar. The company has been heavily criticized for failing to honestly communicate the depth of its breach or disclose the fraudulent certificates to browser makers so they could block them.

In addition to the intelligence agencies, the list of victims to date has included internet giants like Mozilla, Yahoo, Skype, Facebook, Twitter as well as the Tor privacy and anonymizing service and even Microsoft’s Windows Update service, according to Computer World. Certificates issued for Dutch government domains are also believed to have been compromised in the hack.

The Minister of the Interior for the Netherlands said on Saturday that the government could no longer guarantee the security of its websites and urged the public not to log into into them until new certificates could be obtained from other issuing authorities.

DigiNotar acknowledged the breach only after reports began circulating from people in Iran who claimed they were getting browser error messages when they tried to load the Gmail website. Google subsequently confirmed that a fraudulent Google certificate issued to a non-Google entity was operating in the wild, allowing someone to conduct a man-in-the-middle attack to intercept Gmail browsing.

DigiNotar admitted that the hackers who breached its network had obtained certificates for an undisclosed number of domains, but wouldn’t identify the victims. The company has said only that a third-party audit had uncovered a list of certificates the hackers obtained, all of which were subsequently revoked. DigiNotar acknowledged, however, that the auditor had somehow missed the certificate that the hackers had obtained for Google. That certificate was finally revoked last week after Google disclosed its existence in the wild.

Browser makers Google, Mozilla and Microsoft announced this weekend that they would be permanently blocking all digital certificates issued by DigiNotar, suggesting a complete loss of trust in the integrity of its service.

“Based on the findings and decision of the Dutch government, as well as conversations with other browser makers, we have decided to reject all of the Certificate Authorities operated by DigiNotar,” Heather Adkins, a Google information security manager, wrote in a post to the some 300,000 unique IP addresses in Iran may have accessed web sites that used the fraudulent certificate.

“The list of domains and the fact that 99 percent of the users are in Iran suggest that the objective of the hackers is to intercept private communications in Iran,” Fox-IT wrote.

But over the weekend, a hacker who previously claimed credit for breaching Comodo, another certificate authority, earlier in the year claimed responsibility for the DigiNotar breach as well. The hacker, who in the past has identified himself as a 21-year-old Iranian student, claimed he got root access to DigiNotar after obtaining an administrator’s username (Production/Administrator) and password (Pr0d@dm1n). He also claimed to have breached four other certificate authorities, including GlobalSign. Global Sign said in a tweet on Tuesday that it is investigating the claim.

The hacker claimed the attack was retaliation for the Dutch government’s indirect role in the death of 8,000 Serbian Muslims in 1995.

See also:

• Google Certificate Hackers May Have Stolen 200 Others
• Hack Obtains 9 Bogus Certificates for Prominent Websites; Traced to Iran
• Independent Iranian Hacker Claims Responsibility for Comodo Hack

Sony appoints former Homeland Security official as CISO

Sony Corp., which suffered a massive breachof its PlayStation Network earlier this year, is hiring a former official at the Department ofHomeland Security to lead its security initiatives.

Philip Reitinger was named the firms new senior vice president and CISO, in a bid to move onafter multiple high-profile attacks targeting Sonys systems. Reitinger served as directorof the National Cybersecurity Center at DHS. He was also deputy undersecretary of the DHSNational Protection and Programs Directorate (NPPD), a role charged with protecting U.S. governmentsystems from domestic and foreign threats.

In an announcement Tuesday, Sony said Reitinger will overseeinformation security, privacy and Internet safety across the company. He will report to thecompanys executive vice president and general counsel.

Reitinger was chief trustworthy infrastructure strategist at Microsoft, prior to taking the jobat DHS. At Microsoft he oversaw coordination between the software giant, its partners andgovernment agencies on cybersecurity issues.

At a security conference in Miami last year, Reitinger told incident response handlers that itwas time to stoptalking about cybersecurity in generalities and begin taking action against cybercriminals.Mechanisms to help researchers and response teams disseminate information about attacks need to bestandardized for a better coordinated response to incidents, Reitinger said.

"We've got a set of manual processes and there's a lack of agility in places," he told attendeesof the 2010 Forum of Incident Response and Security Teams (FIRST) Conference. "We succeed based ongoodwill and hard work of people rather than the innate design of the system."

SonysPlayStation Network was halted for more than a week, disrupting 77 million PSN and Qriocityaccounts. The company learned that hackers stole personal information about PSN users as well asmore than 24 million Sony Online Entertainment user accounts. Stolen information included names,addresses, email addresses, birth dates and account credentials. The breach also included adatabase containing more than 12,000 non-U.S. credit and debit card numbers

Sony executives have apologized for the security lapses and are giving customers free creditmonitoring, a standard move following breaches. But security experts say Sonys breach highlights anumber of lapses and highlights a need for better datasecurity management.

New Java 7 features improve security

Oracle has officially released Java 7, the first major update to the software platform in fiveyears, to include some security updates and several new features.

The new Java 7features include the use of elliptic curve cryptography and the option to switch off weakerencryption schemes. The Java 7 security enhancements also include improvements to Java SecureSocket Extension and TLS communications, which can prevent some potential attacks.

Oracle says the new Java 7 version coexists with the latest Java 6 Update 27 version and is available for download. Oracle stillmakes use of different installers for the 32 and 64-bit versions for all operating systems (Linux,Solaris & Windows).

Writing on the SANS security blog, researcher Raul Siles warned users to disable Java v6.

From a security perspective, if Java 7 is installed (using Windows as the sampleplatform) on a system that already has Java 6 installed, both versions will remain, so if youonly want to run the latest version, ensure you uninstall any previous versions (as we had to do inthe past but with the same major release) and do not leave vulnerable Java 6 releases around, hesaid. Considering Java is one of the most targeted pieces of client software today, be ready forfuture updates on both, Java 6 and Java 7 in your IT environments (perhaps Java 6u28 and Java7u1), and plan in advance  on how to manage them.

 

U.S. Sources Exposed as Unredacted State Department Cables Are Unleashed Online

An encrypted WikiLeaks file containing 251,000 unredacted U.S. State Department cables is now widely available online, along with the passphrase to open it. The release of the documents in raw form, including the names of U.S. informants around the globe, has raised concerns that dozens of people could now be in danger.

The release of the file comes amidst a heated blame fest between WikiLeaks and the Guardian newspaper in London, which let slip the encrypted version of the database and the decryption key respectively. As details surface about how the leak occurred, it appears that both organizations share the blame.

The 1.73-GB file and passphrase were published Thursday on Cryptome, a competing secret-spilling site, after news broke over the last week that the file had been circulating on the internet unnoticed for several months. Wired.com’s keyword search of the file shows that the uncensored cables contain more than 2,000 occurrences of the phrase “strictly protect”, which is used in cables to denote sources of information whose identities diplomats consider confidential.

It’s unclear how the release will affect imprisoned 23-year-old Pfc. Bradley Manning, who’s facing a court-martial for allegedly leaking the database to WikiLeaks last year.

WikiLeaks had given the Guardian access to the file, along with the passphrase, last summer when WikiLeaks founder Julian Assange met with Guardian editor David Leigh.

WikiLeaks, the Guardian and other media outlets have been publishing the cables in dribs and drabs since last November, after carefully removing the names of most informants. The full database of cables was to have been released piecemeal through Nov. 29 of this year. But last Friday, as news of the leaked file and passphrase was made public, WikiLeaks suddenly began publishing a torrent of cables from the database. It has so far published about 144,000 cables, most of them unclassified. The Associated Press found the names of 90 confidential U.S. sources, including human rights workers laboring under totalitarian regimes, named in that subset of cables.

WikiLeaks said in a statement that it “advanced its regular publication schedule, to get as much of the material as possible into the hands of journalists and human rights lawyers who need it,” before information about the file and passphrase was widely published and repressive regimes sifted through the cables. WikiLeaks has been soliciting votes from the public on whether people agree or disagree that all 250,000 of the cables should be released in raw, unredacted form.

The popular vote favors release, and WikiLeaks has hinted on Twitter its intention to publish. But this time third parties have overtaken the secret-spilling site, and the file is already easily found elsewhere.

WikiLeaks blames the Guardian for disclosing the password in a book it published earlier this year about its WikiLeaks collaboration. WikiLeaks called the Guardian’s action “gross negligence or malice.” “The Guardian disclosure is a violation of the confidentiality agreement between WikiLeaks and Alan Rusbridger, editor-in-chief of the Guardian, signed July 30, 2010,” the group said in a lengthy statement.

The Guardian has downplayed its role in the debacle, while simultaneously revealing a lack of security savvy at the dawn of its relationship with WikiLeaks. The paper notes that although the Guardian’s book did reveal the passphrase, it did not reveal the location of the file, and that Assange had told the paper that “it was a temporary password which would expire and be deleted in a matter of hours. It was a meaningless piece of information to anyone except the person(s) who created the database.”

“No concerns were expressed when the book was published, and if anyone at WikiLeaks had thought this compromised security, they have had seven months to remove the files,” the paper went on to say. “That they didn’t do so clearly shows the problem was not caused by the Guardian’s book.”

Crypto keys, however, last forever, and even if WikiLeaks hadn’t blundered in its handling of the encrypted file, the Guardian clearly should have treated the key as highly sensitive for the foreseeable future.

The fracas heated up last Friday when an editor for the German news weekly Der Freitag revealed that his publication had found the uncensored cables in a 1.73-GB password-protected file named cables.csv” that was available on the internet, and that the password had inadvertently been published online.

WikiLeaks revealed on Wednesday that the passphrase had indeed been published in a book written by Leigh. In the book, Leigh wrote that during the paper’s meeting with Assange in Belgium last year, Assange had given him the passphrase, in part in writing, and in part orally.

Assange had told the paper that the file, which was placed in a subdirectory on a WikiLeaks server, would remain online only a short time, after which it would be removed. Assange, however, apparently never removed the file, and it later found its way into the hands of the organization’s former spokesman, Daniel Domscheit-Berg, and then back to WikiLeaks, after which it wound up on BitTorrent as part of a large archive of WikiLeaks files, which could be downloaded by anyone.

See also:

• WikiLeaks Springs a Leak: Full Database of Diplomatic Cables Appears Online

Survey: APT attacks a top concern, but many firms fail to enforce policies

IT and security professionals are fearful of targetedattacks against their company, but many are failing to put enough safeguards in place to defendagainst them, according to a new survey conducted by Waltham, Mass.-based whitelisting vendor,Bit9.

The Bit9 Endpoint Security Survey polled 765 IT and security professionals in the U.S., Canadaand Europe. More than half the respondents (60%), claimed their main concern was being attacked bycybercriminals that use tactics similar to those used in the RSASecurID breach. Insider threats came in second.

With all the hacks this year, including Sony,which impacted millions of users and is considered to be among the largest breaches ever, it isinteresting that the hacking method that concerns executives the most by a wide margin are theadvanced persistent threat attacks, said Dan Brown, director of security research at Bit9. Thatshows just how serious APTattacks have become and how much damage they can cause enterprises and governmentagencies.

Although security professionals claim they are most concerned about targeted attacks, 50% of thecompanies surveyed said they rely on the honor system for their employees to follow written policyto control and prevent unauthorized software, rather than enforcing it. The survey also found that51% of the companies allow their users to download and install applications.

Brown said this isnt an effective way to protect against downloading software. Basically,companies are telling us they are most worried someone will break down their front door, but theyhavent taken the time to lock it either, He said

In the RSA SecurID breach, attackers used a spear phishing attack on employees at RSA, sendingan email that appeared to be from a co-worker. The attack tricked at least one employee to open anemail attachment, executing malware that targeted an Adobe Flash zero-day vulnerability.

 Normally, when people think about hacks and IT security breaches, they think of theseelaborate James Bond type of attacks involving deception, cracking encrypted codes, usingfingerprints or retinal scans, and other wild methods, explained Brown. The RSA breach used avery simple attack plan that masked an elaborate scam. This attack showed IT executives they needto protect every endpoint or they could end up the next company that is hacked.

Nineteen percent of those surveyed claimed their companies network had crashed due to unusualsoftware on their endpoints, and  89% said the network had been down for less than two hourswhile 13% said it was down for longer than one business day.

According to Brown, not every enterprise will be the target of a malicious attack, but companiesshould still employ precautions, and according to the survey, not all businesses are doing that.The survey found that 74% of the companies allow software to be downloaded only if its approved bythe business, while 17% allow software downloads.

In addition, 79% of respondents claimed their company allows employees to connect removablestorage devices, such as USBs, to their work computers.

So how should enterprises respond to alleviate these threats that theyre facing? According toBrown, by using layered defenses or defense in depth. Its a little bit of motherhood and applepie, but executives need to be aware there are newer attacks and methods out there and they need toprotect against them with new security layers.

PCI tokenization: Vendors need to iron out differences, expert says

The PCI Council has issued its long-awaited guidance supplement addressing the use of tokenizationtechnology to eliminate primary account numbers from merchant systems. Despite the release ofthe report, the tokenization market continues to be plagued with a number of problems, according toUlf Mattsson, chief technology officer of Stamford, Conn.-based tokenization vendor Protegrity Inc.Mattsson said a myriad of token formats have created vendor lock-in, limiting the ability ofmerchants to change systems or use multiple payment processors. Other tokenization systems havescalability issues and cant work with other forms of sensitive data types, Mattsson said.

There was a lot of turmoil that delayed the finalizing of the document.

Ulf Mattson, CTO, Protegrity Inc.

Mattsson, a member of the PCI tokenization special interest group, which helped develop therecent PCI tokenization guidance document, PCITokenization Guidelines, (.pdf) said it took about a year to work out the details of thedocument. Tokenization vendors have long disagreed on a number of details including the supportedtoken formats and the methods used to create tokens. In this interview, Mattson explains that thePCI guidance document is a good first step, but the industry needs to iron out longstandingdisagreements before the technology is wholeheartedly adopted.

How significant is the new PCI guidance document on tokenization?

Ulf Mattsson: I think its much needed. Its a little bit late. Its the first acknowledgementfrom the council about tokenization and its only the first step on the path of validatingtokenization from a PCI point of view. We need to add a lot more steps beyond this supplement.Unfortunately, the supplement contains a lot of disclaimers. The council and the major card brandsadded the disclaimers at the last minute. The council also has no plans to create any certificationat this point.

How difficult was it to come up with a working document?

Mattsson: It was very difficult, even though Visa was able to put out recommendations about ayear ago. The council said it wanted to take the lead because there were a lot of disagreements.There was a lot of turmoil that delayed the finalizing of the document.

What is missing from the document?

Mattsson: I think in the final stages a lot of controversial issues were left out.  [Theguidance document] is starting to give some weight to the technology, but it adds more questionsthan we had before. I think its missing the vendor lock-in issue. When you start to dotokenization its much harder to replace tokens. For example, if you do tokenization with anoutsourcing partner or a payment gateway, you are really stuck with that partner.  Thesetokens are highly customized and theyre all over your applications and databases. They aresometimes deeply integrated into your applications because some applications will not tolerate thetoken format, so there has to be some customization and translation. There should be some wordsabout that lock-in aspect in the document.

Why havent the vendors in this market been able to come together and create an industrystandard to enable companies to change token providers?

Mattsson: Im working in the ASC X9 standards body thatis now addressing tokenization. Were moving in the right direction, but some vendors are pushingtheir model. For example, one vendor is using a form of encryption and they call it tokenization.Thats an extreme example, but weve seen progress. Both Visas best practices and the PCICouncils supplement clearly said is not an acceptable way of generating a token if you want to beout of scope. The council said that is simply an encrypted [primary account number] and anencrypted PAN is within the PCI scope.

What are some other issues holding up standards?

Mattsson: There are other controversial points about outsourcing tokenization. Some vendors arearguing that if you outsource your tokenization, you really have a vendor lock-in situation. So howcan you do this if you need two payment gateways? How can you deal with switching from one paymentsolution provider to another? Larger merchants want to have that function in-house. If we look atupcoming data breach regulations, its not just about the PAN. This is an issue about PII data, PHIdata and cardholder data and they have similar needs to tokenize many of those data types. The onlymodel that will work for them is to have the tokenization server in-house.

The guidance document says theres a possibility to reduce scope with tokenization, but, asyou say, there are a lot of disclaimers. How do you adequately segment components from thetokenization system and the cardholder data environment?

Mattsson: If you take the disclaimers out for a second you see that simply your token serverneeds to be segmented. If you have an in-house tokenization server, you need to put that into aseparate network segment. That is how many QSAs are looking at what is in scope and out ofscope.

Apache DDoS vulnerability requires immediate update to avoid threat

You can legitimately ask for hundreds of very large overlapping parts of a file ina single request. ... A relatively modest number of requests can tie a server's CPU and memory inknots.

Mark Stockley, Web Consultant, Sophos

A new version of the Apache open source Web server, which runs 65% of the worlds websites, hasbeen issued to disable a vulnerability that exposed it to a potential distributed denial-of-service(DDoS) attack.

In an Aug. 31 announcement,, the Apache Software Foundation and the Apache HTTP Server Project said they had released version2.2.20 of the Apache HTTP Server in order to fix the flaw, identified last week. We consider thisrelease to be the best version of Apache available, and encourage users of all prior versions toupgrade, the announcement said.

The new version was produced quickly because a tool that exploits the vulnerability(CVE-2011-31092 at cve.mitre.org) was identified in the wild.

Sophos Web Consultant Mark Stockley wrote on the Sophos Labs Naked Security blog that thevulnerability would allow attackers to mount an ApacheDDoS attack without having masses of computing firepower at their disposal.

The vulnerability can be exploited by a feature in Web servers that allows users to pause andresume their downloads. As Stockley described it: You can legitimately ask for hundreds ofvery large overlapping parts of a file in a single request. Enough parts that a relativelymodest number of requests can tie a server's CPU and memory in knots.

He noted this is partly due to a weakness in the HTTP protocol, meaning other Web servers mightalso be vulnerable.

The new version of Apache reduces the amount of memory used by range requests, and, if the totalbytes of a file requested exceed the total file size, httpd (the Apache HTTP daemon that monitorsincoming requests) will return the entire file.

Network administrators are strongly advised to update their systems immediately. Also writing onthe Sophos blog, Senior Security Advisor Chester Wisniewski observed: Many Linux and Unixadministrators set and forget their installations and never bother to look after their servers.The Apache team should be applauded for testing and releasing an important security fix so quickly.Now it is up to you, the IT administrators, who are using Apache to follow through and apply thesefixes.

Browser makers block rogue SSL certificate

Hackers have acquired a digital certificate from a certificate authority enabling them to issuefraudulent public key certificate requests to a number of domains, including websites owned bysearch engine giant Google.

The certificate breach at Dutch certificate authority, DigiNotar, a subsidiary of VASCO DataSecurity International Inc., gave the cybercriminals the ability to use a rogueSSL certificate to hijack Gmail accounts and spoof secure websites that use SSL and EVSSLdigital certificates for security and to prove their legitimacy to users. The breach took placeJuly 19. In a statement issued by VASCO, the company said it thought it had revokedall fraudulent certificates.

Recently, it was discovered that at least one fraudulent certificate had not been revoked atthe time, the company said.  After being notified by Dutch government organization Govcert,DigiNotar took immediate action and revoked the fraudulent certificate.

The attack was targeted at the systems DigiNotar uses to issue its digital certificates. Thecertificate authority is temporarily suspending the sale of its SSL and EVSSL certificates untilthe conclusion of additional security audits. VASCO said the systems that run its strongauthentication business were not affected by the breach. Details of the stolen certificate wereposted to a public forum last Saturday.

On Monday Google responded to the rogue certificate, claiming it had disabledthe DigiNotar certificate authority in Chrome. The company said the certificate primarilyaffects people in Iran. Mozilla has also disabled support of the certificate.

This means Chrome and Firefox users will receive alerts if they try to visit websites that useDigiNotar certificates, wrote Heather Adkins, an information security manager at Google in theGoogle Online Security blog. To help deter unwanted surveillance, we recommend users, especiallythose in Iran, keep their Web browsers and operating systems up to date and pay attention to Webbrowser security warnings.

Microsoft issued an advisory Monday, announcing it had removedthe DigiNotar root certificate from the list of trusted root certificates for users of WindowsVista and Windows 7.

The certificate potentially affects Internet users attempting to access websites belonging toGoogle, wrote Dave Forstrom, director of Microsoft Trustworthy Computing in the Microsoft SecurityResponse Center blog.  A fraudulent certificate may be used to spoof Web content, performphishing attacks or perform man-in-the-middle attacks against end users.

Attackers have targeted certificate authorities in the past. In March, hackersstole certificates from Comodo Inc. after they penetrated the systems of one of its partnerregistration authorities.The breach resulted in nine fraudulent certificates issued to seven Webdomains, including search engine giants Google and Yahoo. An Iranian hacker claimed responsibilityfor stealing the SSL certificates. Comodo said at no time were any Comodo root keys, intermediateCAs or secure hardware compromised.

U.S. Sources Exposed as Unredacted State Department Cables Are Unleashed Online

An encrypted WikiLeaks file containing 251,000 unredacted U.S. State Department cables is now widely available online, along with the passphrase to open it. The release of the documents in raw form, including the names of U.S. informants around the globe, has raised concerns that dozens of people could now be in danger.

The release of the file comes amidst a heated blame fest between WikiLeaks and the Guardian newspaper in London, which let slip the encrypted version of the database and the decryption key respectively. As details surface about how the leak occurred, it appears that both organizations share the blame.

The 1.73-GB file and passphrase were published Thursday on Cryptome, a competing secret-spilling site, after news broke over the last week that the file had been circulating on the internet unnoticed for several months. Wired.com’s keyword search of the file shows that the uncensored cables contain more than 2,000 occurrences of the phrase “strictly protect”, which is used in cables to denote sources of information whose identities diplomats consider confidential.

It’s unclear how the release will affect imprisoned 23-year-old Pfc. Bradley Manning, who’s facing a court-martial for allegedly leaking the database to WikiLeaks last year.

WikiLeaks had given the Guardian access to the file, along with the passphrase, last summer when WikiLeaks founder Julian Assange met with Guardian editor David Leigh.

WikiLeaks, the Guardian and other media outlets have been publishing the cables in dribs and drabs since last November, after carefully removing the names of most informants. The full database of cables was to have been released piecemeal through Nov. 29 of this year. But last Friday, as news of the leaked file and passphrase was made public, WikiLeaks suddenly began publishing a torrent of cables from the database. It has so far published about 144,000 cables, most of them unclassified. The Associated Press found the names of 90 confidential U.S. sources, including human rights workers laboring under totalitarian regimes, named in that subset of cables.

WikiLeaks said in a statement that it “advanced its regular publication schedule, to get as much of the material as possible into the hands of journalists and human rights lawyers who need it,” before information about the file and passphrase was widely published and repressive regimes sifted through the cables. WikiLeaks has been soliciting votes from the public on whether people agree or disagree that all 250,000 of the cables should be released in raw, unredacted form.

The popular vote favors release, and WikiLeaks has hinted on Twitter its intention to publish. But this time third parties have overtaken the secret-spilling site, and the file is already easily found elsewhere.

WikiLeaks blames the Guardian for disclosing the password in a book it published earlier this year about its WikiLeaks collaboration. WikiLeaks called the Guardian’s action “gross negligence or malice.” “The Guardian disclosure is a violation of the confidentiality agreement between WikiLeaks and Alan Rusbridger, editor-in-chief of the Guardian, signed July 30, 2010,” the group said in a lengthy statement.

The Guardian has downplayed its role in the debacle, while simultaneously revealing a lack of security savvy at the dawn of its relationship with WikiLeaks. The paper notes that although the Guardian’s book did reveal the passphrase, it did not reveal the location of the file, and that Assange had told the paper that “it was a temporary password which would expire and be deleted in a matter of hours. It was a meaningless piece of information to anyone except the person(s) who created the database.”

“No concerns were expressed when the book was published, and if anyone at WikiLeaks had thought this compromised security, they have had seven months to remove the files,” the paper went on to say. “That they didn’t do so clearly shows the problem was not caused by the Guardian’s book.”

Crypto keys, however, last forever, and even if WikiLeaks hadn’t blundered in its handling of the encrypted file, the Guardian clearly should have treated the key as highly sensitive for the foreseeable future.

The fracas heated up last Friday when an editor for the German news weekly Der Freitag revealed that his publication had found the uncensored cables in a 1.73-GB password-protected file named cables.csv” that was available on the internet, and that the password had inadvertently been published online.

WikiLeaks revealed on Wednesday that the passphrase had indeed been published in a book written by Leigh. In the book, Leigh wrote that during the paper’s meeting with Assange in Belgium last year, Assange had given him the passphrase, in part in writing, and in part orally.

Assange had told the paper that the file, which was placed in a subdirectory on a WikiLeaks server, would remain online only a short time, after which it would be removed. Assange, however, apparently never removed the file, and it later found its way into the hands of the organization’s former spokesman, Daniel Domscheit-Berg, and then back to WikiLeaks, after which it wound up on BitTorrent as part of a large archive of WikiLeaks files, which could be downloaded by anyone.

See also:

• WikiLeaks Springs a Leak: Full Database of Diplomatic Cables Appears Online

Morto worm, an old-school Internet worm, spreading via RDP

Various Internet security firms report a new Internet worm is spreading in the wild and takingadvantage of weak passwords on Windows systems, but its exploiting a rarely seen propagationmethod.

First reported Sunday, the Mortoworm or Win32/Morto appears to be an old-school Internet worm, a rarity in recent years whenTrojans and bots make up the majority of new malcode samples.

According to multiple reports, Morto infects Windows workstations and servers, but spreads viathe Windows Remote Desktop Protocol (RDP), an element of the Windows Remote Desktop Connectionservice that allows a Windows PC or server to be controlled remotely.

Once a machine gets infected, the Morto worm starts scanning the local network for machinesthat have Remote Desktop Connection enabled, wrote F-Secure Corp. Chief Research Officer MikkoHypponen in a blog post. This creates a lot of traffic for port 3389/TCP, which is the RDPport.

If it finds such a machine, according to Hypponen, the worm attempts a brute-force login as anadministrator using a series of common passwords. Upon successful login, the worm copies itself tothe new machine, terminates processes associated with local security applications and continues itspropagation attempts. Hypponen also wrote that Morto can be controlled remotely via severalservers, including jaifr.com and qfsl.net.

Microsoft confirmed the existence of the worm in a TechNetblog post Sunday, but it remains unclear which versions of Windows may be vulnerable and theextent to which it is spreading successfully.

Marc Maiffret, CTO of eEye Digital Security, wroteon his companys blog that the Morto worm reminds him of the old days of CodeRed, Slammer,Sasser, Blaster and others. According to Maiffret, companies can avoid infection by disabling RDPaccess directly from the Internet, using strong passwords and making a registry key change so RDPuses non-standard network ports.

One would think that in 2011 such a basic attack would not have much legs, Maiffret wrote,but it seems that antivirus companies and SANS are seeing an increase in RDP networktraffic with the most likely culprit being that of Morto infecting systems via RDP Windows accountbrute-forcing.

In its TechNet post, Microsoft also advised the use of strong passwords, which should include 14characters or more, and have a variety of letters, punctuations, symbols and numbers.

Apache DDoS vulnerability requires immediate update to avoid threat

You can legitimately ask for hundreds of very large overlapping parts of a file ina single request. ... A relatively modest number of requests can tie a server's CPU and memory inknots.

Mark Stockley, Web Consultant, Sophos

A new version of the Apache open source Web server, which runs 65% of the worlds websites, hasbeen issued to disable a vulnerability that exposed it to a potential distributed denial-of-service(DDoS) attack.

In an Aug. 31 announcement,, the Apache Software Foundation and the Apache HTTP Server Project said they had released version2.2.20 of the Apache HTTP Server in order to fix the flaw, identified last week. We consider thisrelease to be the best version of Apache available, and encourage users of all prior versions toupgrade, the announcement said.

The new version was produced quickly because a tool that exploits the vulnerability(CVE-2011-31092 at cve.mitre.org) was identified in the wild.

Sophos Web Consultant Mark Stockley wrote on the Sophos Labs Naked Security blog that thevulnerability would allow attackers to mount an ApacheDDoS attack without having masses of computing firepower at their disposal.

The vulnerability can be exploited by a feature in Web servers that allows users to pause andresume their downloads. As Stockley described it: You can legitimately ask for hundreds ofvery large overlapping parts of a file in a single request. Enough parts that a relativelymodest number of requests can tie a server's CPU and memory in knots.

He noted this is partly due to a weakness in the HTTP protocol, meaning other Web servers mightalso be vulnerable.

The new version of Apache reduces the amount of memory used by range requests, and, if the totalbytes of a file requested exceed the total file size, httpd (the Apache HTTP daemon that monitorsincoming requests) will return the entire file.

Network administrators are strongly advised to update their systems immediately. Also writing onthe Sophos blog, Senior Security Advisor Chester Wisniewski observed: Many Linux and Unixadministrators set and forget their installations and never bother to look after their servers.The Apache team should be applauded for testing and releasing an important security fix so quickly.Now it is up to you, the IT administrators, who are using Apache to follow through and apply thesefixes.

Browser makers block rogue SSL certificate

Hackers have acquired a digital certificate from a certificate authority enabling them to issuefraudulent public key certificate requests to a number of domains, including websites owned bysearch engine giant Google.

The certificate breach at Dutch certificate authority, DigiNotar, a subsidiary of VASCO DataSecurity International Inc., gave the cybercriminals the ability to use a rogueSSL certificate to hijack Gmail accounts and spoof secure websites that use SSL and EVSSLdigital certificates for security and to prove their legitimacy to users. The breach took placeJuly 19. In a statement issued by VASCO, the company said it thought it had revokedall fraudulent certificates.

Recently, it was discovered that at least one fraudulent certificate had not been revoked atthe time, the company said.  After being notified by Dutch government organization Govcert,DigiNotar took immediate action and revoked the fraudulent certificate.

The attack was targeted at the systems DigiNotar uses to issue its digital certificates. Thecertificate authority is temporarily suspending the sale of its SSL and EVSSL certificates untilthe conclusion of additional security audits. VASCO said the systems that run its strongauthentication business were not affected by the breach. Details of the stolen certificate wereposted to a public forum last Saturday.

On Monday Google responded to the rogue certificate, claiming it had disabledthe DigiNotar certificate authority in Chrome. The company said the certificate primarilyaffects people in Iran. Mozilla has also disabled support of the certificate.

This means Chrome and Firefox users will receive alerts if they try to visit websites that useDigiNotar certificates, wrote Heather Adkins, an information security manager at Google in theGoogle Online Security blog. To help deter unwanted surveillance, we recommend users, especiallythose in Iran, keep their Web browsers and operating systems up to date and pay attention to Webbrowser security warnings.

Microsoft issued an advisory Monday, announcing it had removedthe DigiNotar root certificate from the list of trusted root certificates for users of WindowsVista and Windows 7.

The certificate potentially affects Internet users attempting to access websites belonging toGoogle, wrote Dave Forstrom, director of Microsoft Trustworthy Computing in the Microsoft SecurityResponse Center blog.  A fraudulent certificate may be used to spoof Web content, performphishing attacks or perform man-in-the-middle attacks against end users.

Attackers have targeted certificate authorities in the past. In March, hackersstole certificates from Comodo Inc. after they penetrated the systems of one of its partnerregistration authorities.The breach resulted in nine fraudulent certificates issued to seven Webdomains, including search engine giants Google and Yahoo. An Iranian hacker claimed responsibilityfor stealing the SSL certificates. Comodo said at no time were any Comodo root keys, intermediateCAs or secure hardware compromised.

Survey: APT attacks a top concern, but many firms fail to enforce policies

IT and security professionals are fearful of targetedattacks against their company, but many are failing to put enough safeguards in place to defendagainst them, according to a new survey conducted by Waltham, Mass.-based whitelisting vendor,Bit9.

The Bit9 Endpoint Security Survey polled 765 IT and security professionals in the U.S., Canadaand Europe. More than half the respondents (60%), claimed their main concern was being attacked bycybercriminals that use tactics similar to those used in the RSASecurID breach. Insider threats came in second.

With all the hacks this year, including Sony,which impacted millions of users and is considered to be among the largest breaches ever, it isinteresting that the hacking method that concerns executives the most by a wide margin are theadvanced persistent threat attacks, said Dan Brown, director of security research at Bit9. Thatshows just how serious APTattacks have become and how much damage they can cause enterprises and governmentagencies.

Although security professionals claim they are most concerned about targeted attacks, 50% of thecompanies surveyed said they rely on the honor system for their employees to follow written policyto control and prevent unauthorized software, rather than enforcing it. The survey also found that51% of the companies allow their users to download and install applications.

Brown said this isnt an effective way to protect against downloading software. Basically,companies are telling us they are most worried someone will break down their front door, but theyhavent taken the time to lock it either, He said

In the RSA SecurID breach, attackers used a spear phishing attack on employees at RSA, sendingan email that appeared to be from a co-worker. The attack tricked at least one employee to open anemail attachment, executing malware that targeted an Adobe Flash zero-day vulnerability.

 Normally, when people think about hacks and IT security breaches, they think of theseelaborate James Bond type of attacks involving deception, cracking encrypted codes, usingfingerprints or retinal scans, and other wild methods, explained Brown. The RSA breach used avery simple attack plan that masked an elaborate scam. This attack showed IT executives they needto protect every endpoint or they could end up the next company that is hacked.

Nineteen percent of those surveyed claimed their companies network had crashed due to unusualsoftware on their endpoints, and  89% said the network had been down for less than two hourswhile 13% said it was down for longer than one business day.

According to Brown, not every enterprise will be the target of a malicious attack, but companiesshould still employ precautions, and according to the survey, not all businesses are doing that.The survey found that 74% of the companies allow software to be downloaded only if its approved bythe business, while 17% allow software downloads.

In addition, 79% of respondents claimed their company allows employees to connect removablestorage devices, such as USBs, to their work computers.

So how should enterprises respond to alleviate these threats that theyre facing? According toBrown, by using layered defenses or defense in depth. Its a little bit of motherhood and applepie, but executives need to be aware there are newer attacks and methods out there and they need toprotect against them with new security layers.

PCI tokenization: Vendors need to iron out differences, expert says

The PCI Council has issued its long-awaited guidance supplement addressing the use of tokenizationtechnology to eliminate primary account numbers from merchant systems. Despite the release ofthe report, the tokenization market continues to be plagued with a number of problems, according toUlf Mattsson, chief technology officer of Stamford, Conn.-based tokenization vendor Protegrity Inc.Mattsson said a myriad of token formats have created vendor lock-in, limiting the ability ofmerchants to change systems or use multiple payment processors. Other tokenization systems havescalability issues and cant work with other forms of sensitive data types, Mattsson said.

There was a lot of turmoil that delayed the finalizing of the document.

Ulf Mattson, CTO, Protegrity Inc.

Mattsson, a member of the PCI tokenization special interest group, which helped develop therecent PCI tokenization guidance document, PCITokenization Guidelines, (.pdf) said it took about a year to work out the details of thedocument. Tokenization vendors have long disagreed on a number of details including the supportedtoken formats and the methods used to create tokens. In this interview, Mattson explains that thePCI guidance document is a good first step, but the industry needs to iron out longstandingdisagreements before the technology is wholeheartedly adopted.

How significant is the new PCI guidance document on tokenization?

Ulf Mattsson: I think its much needed. Its a little bit late. Its the first acknowledgementfrom the council about tokenization and its only the first step on the path of validatingtokenization from a PCI point of view. We need to add a lot more steps beyond this supplement.Unfortunately, the supplement contains a lot of disclaimers. The council and the major card brandsadded the disclaimers at the last minute. The council also has no plans to create any certificationat this point.

How difficult was it to come up with a working document?

Mattsson: It was very difficult, even though Visa was able to put out recommendations about ayear ago. The council said it wanted to take the lead because there were a lot of disagreements.There was a lot of turmoil that delayed the finalizing of the document.

What is missing from the document?

Mattsson: I think in the final stages a lot of controversial issues were left out.  [Theguidance document] is starting to give some weight to the technology, but it adds more questionsthan we had before. I think its missing the vendor lock-in issue. When you start to dotokenization its much harder to replace tokens. For example, if you do tokenization with anoutsourcing partner or a payment gateway, you are really stuck with that partner.  Thesetokens are highly customized and theyre all over your applications and databases. They aresometimes deeply integrated into your applications because some applications will not tolerate thetoken format, so there has to be some customization and translation. There should be some wordsabout that lock-in aspect in the document.

Why havent the vendors in this market been able to come together and create an industrystandard to enable companies to change token providers?

Mattsson: Im working in the ASC X9 standards body thatis now addressing tokenization. Were moving in the right direction, but some vendors are pushingtheir model. For example, one vendor is using a form of encryption and they call it tokenization.Thats an extreme example, but weve seen progress. Both Visas best practices and the PCICouncils supplement clearly said is not an acceptable way of generating a token if you want to beout of scope. The council said that is simply an encrypted [primary account number] and anencrypted PAN is within the PCI scope.

What are some other issues holding up standards?

Mattsson: There are other controversial points about outsourcing tokenization. Some vendors arearguing that if you outsource your tokenization, you really have a vendor lock-in situation. So howcan you do this if you need two payment gateways? How can you deal with switching from one paymentsolution provider to another? Larger merchants want to have that function in-house. If we look atupcoming data breach regulations, its not just about the PAN. This is an issue about PII data, PHIdata and cardholder data and they have similar needs to tokenize many of those data types. The onlymodel that will work for them is to have the tokenization server in-house.

The guidance document says theres a possibility to reduce scope with tokenization, but, asyou say, there are a lot of disclaimers. How do you adequately segment components from thetokenization system and the cardholder data environment?

Mattsson: If you take the disclaimers out for a second you see that simply your token serverneeds to be segmented. If you have an in-house tokenization server, you need to put that into aseparate network segment. That is how many QSAs are looking at what is in scope and out ofscope.

U.S. Sources Exposed as Unredacted State Department Cables Are Unleashed Online

An encrypted WikiLeaks file containing some 251,000 unredacted U.S. State Department cables is now widely available online, along with the passphrase to open it. The release of the documents in raw form, with the names of U.S. informants around the globe exposed, has raised concerns that of dozens of people could now be in danger.

The release of the file comes amidst a heated blamefest between WikiLeaks and the Guardian newspaper in London, who let slip the encrypted version of the database and the decryption key respectively. As details about how the leak occurred surface, it appears that both organizations share the blame.

The 1.73-GB file and passphrase were published Thursday on Cryptome, a competing secret-spilling site, after news broke over the last week that they had been circulating on the internet unnoticed for several months. A keyword search through the file by Wired.com shows that the uncensored cables contain over 2,000 occurrences of the phrase “strictly protect”, which is used in cables to denote sources of information whose identities diplomats consider confidential.

It’s unclear how the release will affect imprisoned 23-year-old Pfc. Bradley Manning, who’s facing court martial for allegedly leaking the database to WikiLeaks last year.

WikiLeaks had given the Guardian access to the file, along with the passphrase, last summer when WikiLeaks founder Julian Assange met with Guardian editor David Leigh.

WikiLeaks, the Guardian and other media outlets have been publishing the cables in drips and drabs since last November, after carefully removing the names of most informants. The full database of cables was to have been released piecemeal through November 29 of this year. But on Friday, as news of the file and passphrase were made public, WikiLeaks suddenly began publishing a torrent of cables. It has so far published about 144,000 cables, most of them unclassified. The Associated Press found the names of 90 confidential U.S. sources including human rights workers laboring under totalitarian regimes named in that subset of cables.

WikiLeaks said in a statement that it “advanced its regular publication schedule, to get as much of the material as possible into the hands of journalists and human rights lawyers who need it,” before information about the file and passphrase was widely published and repressive regimes sifted through the cables. WikiLeaks has been soliciting votes from the public on whether people agree or disagree that all 250,000 cables should be released in raw, unredacted form. The popular vote favors release, and WikiLeaks has telegraphed on Twitter its intention to publish. But this time third parties have overtaken the secret-spilling site, and the file is already easily found elsewhere.

WikiLeaks blames the Guardian for disclosing the password, calling the act “gross negligence or malice.” “The Guardian disclosure is a violation of the confidentiality agreement between WikiLeaks and Alan Rusbridger, editor-in-chief of the Guardian, signed July 30, 2010,” the group said in a lengthy statement.

The Guardian has downplayed its role in the debacle, while simultaneously revealing a lack of security savviness at the dawn of its relationship with WikiLeaks. The paper notes that although the Guardian’s book did reveal the passphrase, it did not reveal the location of the file, and that Assange had told the paper that “it was a temporary password which would expire and be deleted in a matter of hours. It was a meaningless piece of information to anyone except the person(s) who created the database.”

“No concerns were expressed when the book was published and if anyone at WikiLeaks had thought this compromised security they have had seven months to remove the files,” the paper said. “That they didn’t do so clearly shows the problem was not caused by the Guardian’s book.”

Crypto keys, though, last forever, and even if WikiLeaks hadn’t blundered in its handling of the encrypted file, the Guardian clearly should have treated the key as highly-sensitive for the foreseeable future.

The fracas heated up last Friday when an editor for the German paper Der Freitag revealed that his paper had found the uncensored cables in a 1.73-GB password-protected file named cables.csv” that was available on the internet, and that the password had inadvertently been published online.

WikiLeaks revealed on Wednesday that the passphrase was indeed published in a book written by Leigh. In the book, Leigh wrote that during the paper’s meeting with Assange in Belgium last year, Assange given him the passphrase, in part in writing, and in part orally.

Assange had told the paper that the file, which was placed in a subdirectory on a WikiLeaks server, would remain online only a short time, after which it would be removed. Assange, however, apparently never removed the file and it later found its way into the hands of the organization’s former spokesman, Daniel Domscheit-Berg, and then back to WikiLeaks, after which it wound up on BitTorrent as part of a large archive of WikiLeaks files.

See also:

• WikiLeaks Springs a Leak: Full Database of Diplomatic Cables Appears Online

PCI tokenization: Vendors need to iron out differences, expert says

The PCI Council has issued its long-awaited guidance supplement addressing the use of tokenizationtechnology to eliminate primary account numbers from merchant systems. Despite the release ofthe report, the tokenization market continues to be plagued with a number of problems, according toUlf Mattsson, chief technology officer of Stamford, Conn.-based tokenization vendor Protegrity Inc.Mattsson said a myriad of token formats have created vendor lock-in, limiting the ability ofmerchants to change systems or use multiple payment processors. Other tokenization systems havescalability issues and cant work with other forms of sensitive data types, Mattsson said.

There was a lot of turmoil that delayed the finalizing of the document.

Ulf Mattson, CTO, Protegrity Inc.

Mattsson, a member of the PCI tokenization special interest group, which helped develop therecent PCI tokenization guidance document, PCITokenization Guidelines, (.pdf) said it took about a year to work out the details of thedocument. Tokenization vendors have long disagreed on a number of details including the supportedtoken formats and the methods used to create tokens. In this interview, Mattson explains that thePCI guidance document is a good first step, but the industry needs to iron out longstandingdisagreements before the technology is wholeheartedly adopted.

How significant is the new PCI guidance document on tokenization?

Ulf Mattsson: I think its much needed. Its a little bit late. Its the first acknowledgementfrom the council about tokenization and its only the first step on the path of validatingtokenization from a PCI point of view. We need to add a lot more steps beyond this supplement.Unfortunately, the supplement contains a lot of disclaimers. The council and the major card brandsadded the disclaimers at the last minute. The council also has no plans to create any certificationat this point.

How difficult was it to come up with a working document?

Mattsson: It was very difficult, even though Visa was able to put out recommendations about ayear ago. The council said it wanted to take the lead because there were a lot of disagreements.There was a lot of turmoil that delayed the finalizing of the document.

What is missing from the document?

Mattsson: I think in the final stages a lot of controversial issues were left out.  [Theguidance document] is starting to give some weight to the technology, but it adds more questionsthan we had before. I think its missing the vendor lock-in issue. When you start to dotokenization its much harder to replace tokens. For example, if you do tokenization with anoutsourcing partner or a payment gateway, you are really stuck with that partner.  Thesetokens are highly customized and theyre all over your applications and databases. They aresometimes deeply integrated into your applications because some applications will not tolerate thetoken format, so there has to be some customization and translation. There should be some wordsabout that lock-in aspect in the document.

Why havent the vendors in this market been able to come together and create an industrystandard to enable companies to change token providers?

Mattsson: Im working in the ASC X9 standards body thatis now addressing tokenization. Were moving in the right direction, but some vendors are pushingtheir model. For example, one vendor is using a form of encryption and they call it tokenization.Thats an extreme example, but weve seen progress. Both Visas best practices and the PCICouncils supplement clearly said is not an acceptable way of generating a token if you want to beout of scope. The council said that is simply an encrypted [primary account number] and anencrypted PAN is within the PCI scope.

What are some other issues holding up standards?

Mattsson: There are other controversial points about outsourcing tokenization. Some vendors arearguing that if you outsource your tokenization, you really have a vendor lock-in situation. So howcan you do this if you need two payment gateways? How can you deal with switching from one paymentsolution provider to another? Larger merchants want to have that function in-house. If we look atupcoming data breach regulations, its not just about the PAN. This is an issue about PII data, PHIdata and cardholder data and they have similar needs to tokenize many of those data types. The onlymodel that will work for them is to have the tokenization server in-house.

The guidance document says theres a possibility to reduce scope with tokenization, but, asyou say, there are a lot of disclaimers. How do you adequately segment components from thetokenization system and the cardholder data environment?

Mattsson: If you take the disclaimers out for a second you see that simply your token serverneeds to be segmented. If you have an in-house tokenization server, you need to put that into aseparate network segment. That is how many QSAs are looking at what is in scope and out ofscope.

Google Certificate Hackers May Have Stolen 200 Others

Hackers who obtained a fraudulent digital certificate for Google may have actually obtained more than 200 digital certificates for other top internet entities such as Mozilla, Yahoo and even the privacy and anonymizing service Tor.

Dutch certificate authority DigiNotar, which was hacked in July, has never acknowledged the number of fraudulent certificates the hackers managed to obtain, nor identified the possible targets other than Google.

But a Dutch security consultant told ComputerWorld this week that “about 200 certificates were generated by the attackers.”

Hans Van de Looy, who spoke with the publication, wouldn’t reveal his source, but the number he cited is close to the number of certificates that Google has since placed on the blacklist for its Chrome web browser. On Monday, Google increased the number of certificates its browser was blacklisting from 10 to 247.

News about the hack at DigiNotar broke this weekend after reports began circulating from people in Iran who claimed they were getting browser error messages when they tried to load the Gmail website. Google subsequently confirmed that a fraudulent Google certificate issued to a non-Google entity was operating in the wild, allowing someone to conduct a man-in-the-middle attack to intercept Gmail browsing.

DigiNotar, which is owned by Illinois-based Vasco Data Security, is one of numerous firms around the world that are authorized to generate security certificates to internet entities. The certificates authenticate web pages using the Secure Socket Layer protocol so that users can trust that their encrypted communication is going to the correct location.

DigiNotar acknowledged on Monday that it discovered the breach back on July 19 and said it had revoked all of the certificates the intruders had managed to obtain. But the Google certificate, which had been generated by the intruders on July 10, managed to slip through DigiNotar’s auditors, raising speculation that the Dutch company missed others as well.

Mozilla, which makes the Firefox browser, has since acknowledged that the attackers managed to obtain a certificate for the secure page hosting addons for its browser.

DigiNotar has been criticized for not disclosing the breach earlier to browser makers or the companies, like Google and Yahoo, who have had their digital certificates commandeered.

Survey: APT attacks a top concern, but many firms fail to enforce policies

IT and security professionals are fearful of targetedattacks against their company, but many are failing to put enough safeguards in place to defendagainst them, according to a new survey conducted by Waltham, Mass.-based whitelisting vendor,Bit9.

The Bit9 Endpoint Security Survey polled 765 IT and security professionals in the U.S., Canadaand Europe. More than half the respondents (60%), claimed their main concern was being attacked bycybercriminals that use tactics similar to those used in the RSASecurID breach. Insider threats came in second.

With all the hacks this year, including Sony,which impacted millions of users and is considered to be among the largest breaches ever, it isinteresting that the hacking method that concerns executives the most by a wide margin are theadvanced persistent threat attacks, said Dan Brown, director of security research at Bit9. Thatshows just how serious APTattacks have become and how much damage they can cause enterprises and governmentagencies.

Although security professionals claim they are most concerned about targeted attacks, 50% of thecompanies surveyed said they rely on the honor system for their employees to follow written policyto control and prevent unauthorized software, rather than enforcing it. The survey also found that51% of the companies allow their users to download and install applications.

Brown said this isnt an effective way to protect against downloading software. Basically,companies are telling us they are most worried someone will break down their front door, but theyhavent taken the time to lock it either, He said

In the RSA SecurID breach, attackers used a spear phishing attack on employees at RSA, sendingan email that appeared to be from a co-worker. The attack tricked at least one employee to open anemail attachment, executing malware that targeted an Adobe Flash zero-day vulnerability.

 Normally, when people think about hacks and IT security breaches, they think of theseelaborate James Bond type of attacks involving deception, cracking encrypted codes, usingfingerprints or retinal scans, and other wild methods, explained Brown. The RSA breach used avery simple attack plan that masked an elaborate scam. This attack showed IT executives they needto protect every endpoint or they could end up the next company that is hacked.

Nineteen percent of those surveyed claimed their companies network had crashed due to unusualsoftware on their endpoints, and  89% said the network had been down for less than two hourswhile 13% said it was down for longer than one business day.

According to Brown, not every enterprise will be the target of a malicious attack, but companiesshould still employ precautions, and according to the survey, not all businesses are doing that.The survey found that 74% of the companies allow software to be downloaded only if its approved bythe business, while 17% allow software downloads.

In addition, 79% of respondents claimed their company allows employees to connect removablestorage devices, such as USBs, to their work computers.

So how should enterprises respond to alleviate these threats that theyre facing? According toBrown, by using layered defenses or defense in depth. Its a little bit of motherhood and applepie, but executives need to be aware there are newer attacks and methods out there and they need toprotect against them with new security layers.

This Week: Appeals Court to Weigh NSA Dragnet Surveillance

Whether the federal government and the nation’s telecommunication companies can be held accountable for allegedly funneling every American’s electronic communication to the National Security Agency without warrants is the subject of oral arguments scheduled for a federal appeals court Wednesday.

At issue is a Jan. 31, 2006 lawsuit, and others that followed, alleging violations of the Fourth Amendment right to be free from warrantless searches and seizures. The cases, about three dozen which will be consolidated into two oral arguments, have been thrown out of court on a variety of grounds, chiefly the government’s claim that the lawsuits would expose state secrets, and a 2008 law that immunized the nation’s telcos from such lawsuits.

Nearly six years later, the merits of the lawsuits have never been addressed. The Electronic Frontier Foundation, which brought the leading cases, appealed, and contends that the litigation should never have been dismissed.

“As far as we know the surveillance is ongoing,” says Cindy Cohn, the EFF’s legal director, who will be arguing before a three-judge panel of the 9th U.S. Circuit Court of Appeals in Seattle. “I think it is tremendously important that Americans not be subject to dragnet surveillance by the government. I think the Fourth Amendment, the right to privacy, is important for this country.”

Threat Level will cover the arguments from the courtroom Wednesday afternoon. The hearing is expected to begin at 2:00 p.m. and last at least two hours.

The Obama administration is set to urge the court to let stand the lower-court decisions dismissing the lawsuits. What’s more, the government contends, litigation against the government and the telcos must die because it threatens to expose government secrets and undermine national security.

“Congress made a legislative policy judgment that, if litigation of this kind were permitted to proceed, firms that may have assisted the nation at a critical time would be improperly burdened and sensitive classified information might be improperly disclosed,” Thomas Bondy, a Justice Department attorney, told the appeals panel in a court brief.

The EFF’s allegations are based in part oninternal AT&T documents allegedly outlining secret rooms in AT&T offices that route internettrafficto the NSA. Every major telecom carrier in the United States is now named in at least one of the surveillance lawsuits for allegedlycooperatingwith the government’s warrantlesssurveillance program.

The Bush administration, and now the Obama administration, have neither admitted nor denied the allegations. Instead, they have declared the issue a state secret — one that would undermine the nations national security if exposed.

U.S. District Judge Vaughn Walker, the San Francisco judge presiding over the litigation, did not agree. The judge had ruled the allegations against the nation’s telcos could proceed.

But a major obstacle stopped the case dead in its tracks, before the merits of the allegations could be litigated, and before the judge could consider ordering a halt to the alleged dragnet, as the EFF is demanding.

That roadblock was an act of Congress, one voted for by then-senator Barack Obama of Illinois, and then signed by President George W. Bush in July 2008. The legislation handed the telcos retroactive immunity from being sued for participating in the surveillance program. That led Judge Walker totoss the case. But EFF contends on appeal that the legislation, which grants the president the power to grant immunity to the telcos, was an unlawful abuse of power.

“It gives a giant scepter to the executive branch to dispense immunity, a civil pardon,” Cohn said. “That’s the constitutional problem.”

That same immunity legislation also approved Bush’s once-secret warrantless eavesdropping, which The New York Times disclosed in December 2005.

EFF and others countered theimmunitylaw by naming the government, instead of the telcos, as the defendant. That promptedObama to invoke the state secretsprivilege– despite havingannounced he would limit his use of that doctrine.Judge Walker wound up dismissing the revised lawsuit as a “general grievance” from the public, and not an actionable claim.

Photo: Mark Klein

See Also:

• After Cheating Scandal, FBI Agents to Be Tested On Surveillance Tactics
• ACLU Study Highlights U.S. Surveillance Society
• Court Says Bush Illegally Wiretapped Two Americans
• Lawmakers Propose Warrant Requirement for GPS Data
• Domestic Surveillance Court Approved All 1,506 Warrant Applications in 2010
• Feds: Privacy Does Not Exist in Public Places
• San Francisco Subway Shuts Cell Service to Foil Protest; Legal Debate Ignites

Browser makers block rogue SSL certificate

Hackers have acquired a digital certificate from a certificate authority enabling them to issuefraudulent public key certificate requests to a number of domains, including websites owned bysearch engine giant Google.

The certificate breach at Dutch certificate authority, DigiNotar, a subsidiary of VASCO DataSecurity International Inc., gave the cybercriminals the ability to use a rogueSSL certificate to hijack Gmail accounts and spoof secure websites that use SSL and EVSSLdigital certificates for security and to prove their legitimacy to users. The breach took placeJuly 19. In a statement issued by VASCO, the company said it thought it had revokedall fraudulent certificates.

Recently, it was discovered that at least one fraudulent certificate had not been revoked atthe time, the company said.  After being notified by Dutch government organization Govcert,DigiNotar took immediate action and revoked the fraudulent certificate.

The attack was targeted at the systems DigiNotar uses to issue its digital certificates. Thecertificate authority is temporarily suspending the sale of its SSL and EVSSL certificates untilthe conclusion of additional security audits. VASCO said the systems that run its strongauthentication business were not affected by the breach. Details of the stolen certificate wereposted to a public forum last Saturday.

On Monday Google responded to the rogue certificate, claiming it had disabledthe DigiNotar certificate authority in Chrome. The company said the certificate primarilyaffects people in Iran. Mozilla has also disabled support of the certificate.

This means Chrome and Firefox users will receive alerts if they try to visit websites that useDigiNotar certificates, wrote Heather Adkins, an information security manager at Google in theGoogle Online Security blog. To help deter unwanted surveillance, we recommend users, especiallythose in Iran, keep their Web browsers and operating systems up to date and pay attention to Webbrowser security warnings.

Microsoft issued an advisory Monday, announcing it had removedthe DigiNotar root certificate from the list of trusted root certificates for users of WindowsVista and Windows 7.

The certificate potentially affects Internet users attempting to access websites belonging toGoogle, wrote Dave Forstrom, director of Microsoft Trustworthy Computing in the Microsoft SecurityResponse Center blog.  A fraudulent certificate may be used to spoof Web content, performphishing attacks or perform man-in-the-middle attacks against end users.

Attackers have targeted certificate authorities in the past. In March, hackersstole certificates from Comodo Inc. after they penetrated the systems of one of its partnerregistration authorities.The breach resulted in nine fraudulent certificates issued to seven Webdomains, including search engine giants Google and Yahoo. An Iranian hacker claimed responsibilityfor stealing the SSL certificates. Comodo said at no time were any Comodo root keys, intermediateCAs or secure hardware compromised.

Couple Can Sue Laptop-Tracking Company for Spying on Sex Chats

An Ohio woman and her boyfriend can sue a laptop-tracking company that recorded their sexually explicit communications in an effort to identify thieves who stole the computer the woman was using.

U.S. District Judge Walter Rice ruled last week against Absolute Software, which provides software and services for tracking stolen computers. Absolute sought a summary judgment in its favor, insisting that one of its theft recovery agents acted properly when he captured sexually explicit images of Susan Clements-Jeffrey communicating via webcam with her boyfriend and passed them to police in an effort to recover the stolen computer.

But the judge found that there were grounds to believe Absolute had gone too far, and that a jury might reasonably decide that it had violated the plaintiffs’ privacy and broken the law. The case raises an important issue about the length that someone can legally go to recover stolen goods.

“It is one thing to cause a stolen computer to report its IP address or its geographical location in an effort to track it down,” Rice wrote in his decision (.pdf). “It is something entirely different to violate federal wiretapping laws by intercepting the electronic communications of the person using the stolen laptop.”

The case revolves around a laptop that Clements-Jeffrey, a substitute teacher, bought from one of her students in 2008.

The laptop belonged to Clark County School District in Ohio, and had been stolen from one of its students in April 2008. Another student at Kiefer Alternative School subsequently purchased the laptop at a bus station for $40, even though he suspected it was stolen, and turned around and offered it to Clements-Jeffrey for $60.

Clements-Jeffrey, who was a long-term substitute teacher at Kiefer, says the student told her his aunt and uncle had given him the laptop, but that he no longer needed it after getting a new one. She asserts she had no idea the computer was stolen.

Clements-Jeffrey, described in court papers as a 52-year-old widow, had recently renewed a romance with her high school sweetheart, Carlton Smith, who lived in Boston. In the course of their courtship, she exchanged sexually explicit email and instant messages with her beau, using the computer she had just purchased.

What she didn’t know was that Clark County School District, which legally owned the laptop, had purchased Absolute’s theft recovery service, which includes the installation of its remote-recovery software LoJack, onto client computers. The system gives Absolute employees remote access to a stolen computer and allows them to record and intercept any data from the machine.

After the school district reported the laptop stolen, Absolute began collecting the IP address from Clements-Jeffrey’s laptop when it connected to the internet.

Ordinarily, the next step would be for Absolute to provide a suspect’s IP address to law enforcement agents, so that they could issue a subpoena to the suspect’s ISP to obtain the user’s name and physical address. But Absolute’s theft officer Kyle Magnus went further and began to remotely intercept e-mail and other electronic communications going to and from Clements-Jeffrey’s machine in real time.

According to court documents, in June 2008 Magnus began recording Clements-Jeffrey’s keystrokes and monitoring her web surfing. At one point, while snooping on Clements-Jeffrey’s webcam communications with her boyfriend, Magnus also captured three screenshots from her laptop monitor, which showed Clements-Jeffrey naked in the webcam images. In one picture, her legs were spread apart.

Magnus subsequently sent the pictures and recorded communications, along with Clements-Jeffrey’s name and contact information, to a police detective. When the police showed up at the plaintiff’s apartment to collect the laptop, they were brandishing the explicit images Magnus had sent them. They then arrested and charged her for receiving stolen property. The charges, however, were dismissed about a week later.

Clements-Jeffrey and her boyfriend, Smith, sued Absolute Software, Kyle Magnus, the city of Springfield, Ohio, and two police officers. The plaintiffs allege that the police violated their Fourth Amendment rights, and that Absolute violated the Electronic Communications Privacy Act and the Stored Communications Act and intentionally invaded their privacy.

The case rests largely on whether Clements-Jeffrey knew the laptop she bought was stolen and whether she and her boyfriend then had a reasonable expectation of privacy.

The defendants moved for summary judgment on grounds that courts have ruled in the past that there is no legitimate expectation of privacy in cases involving known stolen property. They asserted that Clements-Jeffrey should have known the laptop was stolen based in part on the $60 price the seller was asking for it and on the fact that the serial number had been scraped off the bottom of the machine.

Clements-Jeffrey, however, asserted she never noticed the missing serial number and had no reason to doubt the asking price for the two-year-old machine, since the computer had been wiped clean of software before she bought it. She said Absolute had a right to collect her IP address in an effort to track the laptop, but that it broke the law when it intercepted her communications to track her and then passed those images to police. The ECPA statute prohibits intercepting or disclosing the contents of someone’s wire or other electronic communications without their knowledge.

Absolute also insisted it was acting on behalf of its customer, the school district, and therefore was covered under “color of law” and “safe harbor” statutes. The company cited its agreement with the school district, which gives Absolute’s staff “the ability to view and recover any files that are present” on the school’s computers.

But the school district has asserted that it never knew this meant that Absolute would intercept communications that a suspected thief might have with third parties.

The judge ultimately ruled that although Absolute might have had a noble purpose in assisting the school district in recovering its laptop, “a reasonable jury could find that they crossed an impermissible boundary.”

According to Absolute’s web site, it recovers on average 14 laptops a day. Asked if the company’s agents have changed the way they operate in light of the lawsuit, Absolute spokesman Stephen Midgley declined to respond. “Because it’s currently still under legal proceedings, Absolute isn’t commenting on the story at this time,” he said.

Photo: Jim Merithew/Wired

Feds, EFF Clash in Appeals Court Hearing on NSA Spying

Judge Michael Daly Hawkins and Judge M. Margaret McKeown consider whether to reinstate lawsuits accusing the government and telecom companies of siphoning electronic communications to the NSA.

SEATTLE –A three-judge federal appeals court grilled government and civil rights lawyers while entertaining arguments here Wednesday concerning dozens of dismissed lawsuits alleging the National Security Agency illegally vacuumed American’s internet traffic and telephone calls from every major U.S. telecommunication company.

But after nearly three hours of back-and-forth debate that included harsh questioning by the judges of the 9th U.S. Circuit Court of Appeals, it was unclear whether the court would reinstate the cases that allege rampant, warrantless spying. And despite the import of the court’s pending decisions, Judge Harry Pregerson, a President Jimmy Carter appointee and one of the nation’s longest-serving jurists, joked about the subject matter.

“I’m used to electronic surveillance, I live with it every day,” Pregerson said as the roughly three dozen members of the gallery laughed out loud. Pregerson appeared via a live video feed from Los Angeles and was not present here at the William Kenzo Nakamura United States Courthouse.

The 9th U.S. Circuit Court of Appeals, the nation’s largest federal appeals court, is based in San Francisco and covers nine western states, including Alaska, Arizona, California, Hawaii, Idaho, Montana, Nevada, Oregon and Washington. There are 26 active judges on the circuit. The court hears cases with three-judge panels, which are generally chosen at random.

Here are the judges on the National Security Agency wiretapping lawsuits:

Harry Pregerson, 87
Appointed to U.S. District Court in Los Angeles by President Lyndon Johnson
Appointed to 9th Circuit by President Jimmy Carter
Law School: University of California, Berkeley Boalt Hall School of Law, 1950
Based in Los Angeles

Michael Daly Hawkins, 66
Appointed to 9th Circuit by President Bill Clinton
Law School: Arizona State University, 1970
Based in Phoenix

M. Margaret McKeown, 60
Appointed to 9th Circuit by Clinton
Law School: Georgetown University Law Center, 1975
Based in San Diego

One set of cases on appeal, originally brought by the Electronic Frontier Foundation in 2006, targets the nation’s telecoms. The lawsuit alleges the carriers’ alleged complicity in the electronic spying had breached federal wiretapping laws and even their own terms of service agreements with customers.Another round of litigation, brought by the EFF and others, targets the government, accusing federal officials of violating the Fourth Amendment rights of anybody who so-much as sent an e-mail in the years following the September 11, 2001 terrorist attacks. (Some of the lawsuits allege the dragnet began before 9/11.)

“Who was or who was not surveilled, that’s classified … What any particular carrier did or did not do, that’s all classified.”

All the cases were previously dismissed by a San Francisco federal judge on a variety of grounds, including a decision that Americans had no legal standing to bring a challenge against the government. Many of the cases were dismissed under a law passed by Congress in 2008 that authorized the president, in this case Barack Obama, to give retroactive legal immunity to any telephone companies that cooperated with the alleged warrantless surveillance.

“We think this gives too much power to the executive. The Constitution has limits as to how much power,” Cindy Cohn, the EFF’s legal director, told the panel.

Judge Michael Daly Hawkins wondered aloud, “If these plaintiff’s don’t have standing, who would?” Judge M. Margaret McKeown said the “concern” she had was that the suits’ dismissal “cuts off the plaintiffs … from ever pursuing a claim.”

Michael Kellogg, the carriers’ attorney, argued the immunity legislation was the right thing for the nation’s carriers, which could go bankrupt under the weight of defending the accusations in court.

“Congress made a considered decision that it would be unfairif they were subject to potential suits and ruinous liability,” Kellogg said.

Department of Justice Attorney Thomas Bondy urged the panel of judges to abide by Congress’ wishes. He repeated over and again that litigating the allegations would expose national security secrets.

“Who was or who was not surveilled, that’s classified,” he said. “Whatany particular carrier did or did not do, that’s all classified.”

Judge Harry Pregerson, appearing via a live video feed

The EFF, a San Francisco-based civil rights group, believes the surveillance dragnet continues unabated today, in effect granting the government unfettered access to Americans’ private lives. The merits of the allegations have never been litigated, and the EFF is hoping the appeals court reinstates the cases toward that end.

Judge Pregerson asked Department of Justice attorney Thomas Byron III, “What role would the judiciary have if your approach was adopted. We just get out of the way, is that it?”

“We think the district court’s order of dismissal can be upheld,” Byron replied.

The The New York Times first exposed the NSA’s warrantless wiretapping of international phone calls to and from Americans in 2005. A former AT&T technician named Mark Klein later produced internal company documents — first published by Wired.com — suggesting that the NSA was surveilling internet backbone traffic from a secret room at an AT&T switching center in San Francisco, and similar facilities around the country. Klein’s evidence formed the basis of one of EFF’s lawsuits, Hepting v. AT&T.

Department of Justice attorneys Thomas Byron, front, and Thomas Bondy, wait their turn to address the court

First the Bush administration, and then the Obama administration, fiercely fought the lawsuits. Both administrations trotted out the state secrets privilege, a McCarthy-era doctrine that generally allows the government to quash a lawsuit, even if the government is not a defendant, whenever litigating the case could damage national security and expose state secrets. In an rare defeat for the government, in 2008 U.S. District Judge Vaughn Walker of San Francisco allowed the Hepting lawsuit to go forward despite the state secrets claim.

EFF senior staff attorney Kevin Bankston urges the three-judge panel to hold the government accountable

After Walker’s blockbuster decision, Congress passed legislation in 2008 immunizing the telecommunication companies from the lawsuits. The legislation, which then-Sen. Barack Obama voted for, also granted the government the authority to monitor Americans’ telecommunications without warrants if an American was communicating with somebody overseas and suspected of terrorism — effectively legalizing at least one facet of the alleged NSA dragnet.

Because of the immunity legislation, Walker dismissed the case against the carriers in 2009, saying Congress had spoken.

“Congress has manifested its unequivocal intention to create an immunity that will shield the telecommunications company defendants from liability in these actions,” Walker ruled.

The EFF then sued the NSA, which is not protected by the immunity law. Walker dismissed that case, Jewell v. NSA, last year on the grounds that the lawsuit amounted to a “general grievance.”

Kevin Bankston, an EFF senior staff attorney, told the judges that the government should be held accountable, and that Jewell should be reinstated. “We’d love to get to the merits,” he said.

The court did not indicate when it would rule.

Photos: Jon Snyder/Wired.com

WikiLeaks Springs a Leak: Full Database of Diplomatic Cables Appears Online

Julian Assange (Photo: WikiMedia Commons)

For the second time in a year, WikiLeaks has lost control of its full, unredacted cache of a quarter-million U.S. State Department cables — and this time the leaked files are apparently online.

The uncensored cables are contained in a 1.73-GB password-protected file named “cables.csv,” which is reportedly circulating somewhere on the internet, according to Steffen Kraft, editor of the German paper Der Freitag. Kraft announced last week that his paper had found the file, and easily obtained the password to unlock it.

Unlike the cables that WikiLeaks has been publishing piecemeal since last fall, these cables are raw and unredacted, and contain the names of informants and suspected intelligence agents that were blacked out of the official releases. Der Freitag said the documents include the names of suspected agents in Israel, Jordan, Iran and Afghanistan, and noted that interested parties — such as the Iranian government or intelligence agencies — could have already discovered and decrypted the file to uncover the names of informants.

“The story is that a series of lapses, as far as I can see on behalf of WikiLeaks and its affiliates, has led to the possibility a file becoming generally available which it never should have been available,” confirmed former WikiLeaks staffer Herbert Snorrason, of Iceland, who left the organization as part of a staff revolt last year, and is now part of the competing site OpenLeaks.

Information about the exposed file and password was also confirmed by the German newsweekly Der Spiegel. According to that publication, the cables were contained in an encrypted file that WikiLeaks founder Julian Assange had stored on a subdirectory of the organization’s server last year, which wasn’t searchable from the internet by anyone who didn’t already know its location.

Assange had reportedly given the password for the file to an “external contact” to access the file’s contents. With both the file and the password now online, the leak is complete.

“The issue is double: On one hand there is the availability of the encrypted file, and on the other the release of the password to the encrypted file,” Snorrason told Threat Level on Monday. “And those two publications happened separately.

The password leak was done “completely inadvertently,” Snorrason added. He declined to identify the leaker, or the circumstances of the leak, but said it was someone who was with neither WikiLeaks nor OpenLeaks.

Last year, former WikiLeaks spokesman Daniel Domscheit-Berg and another WikiLeaks staffer led a staff revolt at WikiLeaks following a rift with Assange. They finally left the organization and set up OpenLeaks.org. When they left WikiLeaks, they took the contents of the WikiLeaks server with them, which included the encrypted file. Last December, Domscheit-Berg returned most of what he had taken, including the file containing the cables.

Wikileaks supporters subsequently released an archive of the data that Domscheit-Berg had returned, as a public service to provide readers with access to everything WikiLeaks had previously published. But among the documents was the encrypted file containing the cables. Several months later, the person to whom Assange had provided the password somehow made it public online. Der Spiegel doesn’t elaborate on precisely why or how that person published the password, and Snorrason declined to say more, for fear of guiding people to the password.

“It’s not very obvious how the password was made available, and we’re not keen on making it any more obvious how or why it might have been published,” Snorrason said.

Both the encrypted file and password went unnoticed until recently. Der Spiegel implies that Domscheit-Berg or someone else connected to his rival OpenLeaks organization was responsible for calling Der Freitag’s attention to the file and password to make a point that WikiLeaks is unable to properly secure the data it possesses. Domscheit-Berg did not immediately respond to an e-mail query from Threat Level on Monday.

After nine months of slow, steady publication, WikiLeaks abruptly opened the spigot last week on its cable publications, spewing out over 130,000 by Monday afternoon — more than half the total database.

This is not the first time that WikiLeaks has lost control of its database of cables. Last year, as the organization and its media partners were beginning preparations to publish stories related to the cables, a WikiLeaks member gave the database to a freelance reporter, Heather Brooke.

Brooke was not a member of the approved cabal of media outlets that had been given access to the documents and her possession of them threatened to derail the plans that WikiLeaks and its media partners had hammered out for publication. The Guardian newspaper in the U.K. subsequently secured agreement from Brooke that she wouldn’t herself publish any of the cables or stories related to them.

WikiLeaks responded to the leak on Twitter on Monday by writing: “There has been no ‘leak at WikiLeaks’. The issue relates to a mainstream media partner and a malicious individual.”